The Denison Experience versus Partisanship
Most students feel neutral or negatively about Biden’s foreign and domestic economic policies.
They feel better about his performance with student loans.
Students that think more positively about Biden’s performance in one area are more likely to think positively about his performance in another area.
Do student experiences at Denison influence their attitudes?
Or is partisanship more powerful?
Students come to Denison with their own set of perspectives and beliefs.
Is there any room for these perspectives and beliefs to evolve?
Denison is supposed to be a place where students are exposed to new ideas and view points.
If this is true, shouldn’t the length of time students are at Denison or their area of study move the needle on an ideologically charged question (like how well is the current U.S. President doing?).
The student experience:
Class (Freshman, Sophomore, Junior, Senior).
Area of major (Social Sciences, Sciences, Humanities, Arts, Interdisciplinary).
Partisanship:
Self-identified partisanship (from “Strong Democrat” to “Strong Republican”).
Interest in politics (from “Not at all” to “Extremely” interested).
Polarization (absolute difference in support for campus Republicans vs. Democrats).
No classes think Biden has performed terribly well.
Sophomores are generally more positive than other classes.
There’s no evidence that student attitudes systematically change with their year in school.
There are also very few differences by academic divisions.
Interdisciplinary majors are more supportive than others in terms of Biden’s foreign and domestic economic policies.
Everyone, except majors in the social sciences, thinks Biden has been doing better on student loans (with art students the most supportive).
We see the biggest differences due to partisanship.
Foreign policy: Democrats are at best neutral while Republicans are explicitly negative.
Domestic economic policy: Democrats are mildly positive while Republicans are strongly negative.
Student loans: Democrats are clearly positive while Republicans lean negative (but not with a high degree of precision).
Affective polarization also influences student attitudes.
This is especially true for partisans.
The more polarized students are, the further apart their attitudes become.
Interest in politics matters, too, and it also is moderated by partisanship.
The more interested in politics partisans are, the more their attitudes diverge.
Permutation importance metrics from random forest models show the absolute change in the quality of predictions by simulating what would happen if we “broke” the relationship between a predictor and an outcome.1
Higher values mean a variable is really important for making predictions.
Political factors are much more important that factors related to student’s time at Denison.
Partisanship dominates over students’ Denison experience.
Partisanship is by far the most important factor explaining how students evaluate Biden’s policies.
Meanwhile, year in school and area of study have a smaller impact.
Denison University | Data for Political Research
Predictive models done with random forest regressions using the {ranger} R package.